I needed a rope, so I created one by using structural panels and connected them using rotators with the bare look, with range of rotation restricted to something like 50 degrees. I connected the rotators alternatingly vertically and horizontally such that the rope can curl both horizontally and vertically. After connecting them I had to rotate the connected panel to the right orientation and I moved it a little such that the panels are in line and such that the panels can rotate without touching each other (the small thick end points of the panels are only there in the editor).
Now, in the simulation I pulled on the rope from the left side, and on the right side there is some small weight but it is not fixed, the whole rope can move freely. As you can see, the rope stretches: large gaps appear between the panels and somehow this happens only at the half of the rotators that are in the same orientation.
I see no way to fix this with the tinker options. Sidenote: I used the tinker panel to scale these parts down and made them weight much less or nothing, but other than that I didn't do anything strange.
I don't know what causes this. I tried to fix it by increasing the rigidness factor (or whatever it was called, from the tinker submenu) from 1 all the way up to 50 for these parts but it doesn't make any difference.
@AndrewGarrison thanks for the answer, I checked the configuration I used and I tried realistic scaling (tinker size by factor x, then scale mass by factor x^2). The structural panels are on 25% scale and 5% mass, which is nicely lightweight. The rotators in between are also on 25% scale but I had them on 0% mass because when using 5% mass then they are still 2.5 kg each, which makes the mass ridiculously high for what is basically just a rope. When setting the mass of all the rotators to 5%, the problem seems to be solved, so that is nice.
However, two points remain that I want to make:
1. The fact that the gap occurred on half of the rotators is still kinda strange to me, the orientation of the rotator should not really matter for how it behaves in the simulation of the situation I showed. I understand that zero mass is not garantueed to work correctly, but at least it could point to some inconsistency in the simulation code.
2. I would suggest changing the sliders in the tinker panels to decimal number input fields between 0 and 1. Or if you want to keep the slider, then finer steps around zero, so a non-linear scaling.